{"id":"a2737eb7-7c6f-484d-a5bf-a658d2268985","arxiv_id":"2507.23466","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":9,"one_line_summary":"With adaptive-optics pre-compensation and optimistic detector parameters, GEO-satellite TF-QKD and MP-QKD can reach hundreds of bit/s with 1 m ground telescopes, falling to 17 bit/s with current space detectors and positive key rates at 20 cm only for idealized detectors.","lead":"This paper simulates twin-field and mode-pairing quantum key distribution from a geostationary satellite to ground stations, reporting secret key rates up to a few hundred bits per second in the best case. It suggests such links could work with ground telescopes as small as 20 cm, but only if space-grade detectors reach ground performance.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The positive-rate result hinges on the simulated MMSE adaptive-optics gain; the paper states that without MMSE no positive key rate is obtained, yet the MMSE model is not independently validated.","rationale":"The reader's weakest assumption is the same as mine: the advanced MMSE adaptive-optics pre-compensation. I agree with that identification and do not find an internal inconsistency that would change the verdict. The strongest independent support is the demonstrated AO pre-compensation in Refs [23,24], but those demonstrations correspond to SoA-level correction and do not validate the MMSE gain. The paper's own admission that SoA correction yields zero key rate for all apertures is the clearest evidence that the MMSE model is load-bearing. I therefore keep the reader's CONDITIONAL verdict, with the condition being an independent validation of the MMSE coupling statistics and a sensitivity analysis around the positive-rate threshold.","tokens_in":19000,"tokens_out":11782,"duration_ms":140196,"concrete_test":"Use the authors' rate code to find the mean-ηturb value at which R crosses zero for DOGS=1 m, then run an independent end-to-end wave-optics simulation of the MMSE pre-compensated GEO uplink that includes finite AO loop delay, wavefront-sensor read noise, and the actual point-ahead covariance, and feed the resulting ηturb distribution into the TF/MP key-rate calculation. If the independent mean coupling (or the simulated tail of low ηturb) falls below that threshold, the positive-rate claim fails.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim (abstract: 'few hundred bit/s' for TF and MP, and feasibility down to 20 cm) is carried by the MMSE pre-compensation channel model. The authors themselves state in Section IV C 1 that under state-of-the-art (SoA) correction 'we did not obtain a positive key rate for any aperture diameter under these conditions.' Thus every positive rate in the paper exists only because the MMSE estimator is assumed to perform as modeled in Refs [32,62], which are simulations by the same group. The margin is not large: Table II gives mean ηturb at DOGS=1 m of 0.56 (MMSE) versus 0.40 (SoA), roughly 1.5 dB in the mean, and at 20 cm the means are 0.73 versus 0.72. Because the QKD rate is a threshold function of the tail of the PDTE, a modest error in the simulated MMSE tail can move the result from 260 bit/s to zero. The covariance model in Eqs. (A14)-(A15) assumes known phase/amplitude statistics, no wavefront-sensor noise, no temporal error, and exact PAA covariance; none of these assumptions is stress-tested. This makes the headline feasibility claim conditional on an unvalidated, self-referenced AO model.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper analyzes the feasibility of untrusted-node satellite QKD using twin-field (TF) and mode-pairing (MP) protocols with a geostationary satellite as the central untrusted node. It develops an end-to-end channel model that includes atmospheric turbulence with adaptive optics (SoA and MMSE pre-compensation), satellite pointing jitter, geometrical/absorption/system losses, and detector dark counts and efficiency. The authors simulate the probability distribution of the end-to-end transmission efficiency and compute asymptotic secret key rates for TF-QKD and MP-QKD for ground telescope apertures from 20 cm to 1 m, under three detector scenarios (optimistic, pessimistic/space-demonstrated, and idealized). They find, in the optimistic scenario, maximum rates around 260 bit/s (TF) and 180 bit/s (MP) at 1 m apertures, and positive rates at 20 cm only for the idealized detector parameters; with only state-of-the-art AO correction no positive key rate is found for any aperture.","tokens_in":19338,"tokens_out":10007,"duration_ms":98365,"significance":"This is one of the first detailed end-to-end simulations of an untrusted-node GEO satellite QKD architecture with small-aperture ground telescopes. The work is thorough in its channel modeling: it uses the reciprocity principle for pre-compensated uplinks, a pseudo-analytical turbulence model, a realistic jitter model, and the established asymmetric TF and MP-QKD security formulas. It also compares multiple detector scenarios, which is useful for assessing technology roadmaps. The key-rate predictions are quantitative and falsifiable. However, the central feasibility claim is conditional on the MMSE adaptive-optics model (not independently validated here) and on the optimistic detector scenario for the headline numbers; these conditions must be clearly communicated.","major_comments":[{"comment":"The abstract's central claim of 'a few hundred bit/s for both TF and MP-QKD' 'considering realistic detectors' is not supported by the detector scenario that corresponds to currently demonstrated space technology. With the space-qualified detector parameters from [49] (pd=4e-8, etaD=50%), the paper reports positive rates only for MP-QKD with a 1 m OGS aperture, at 17 bit/s (Section IV C 3), and no positive rate for TF-QKD. The few-hundred-bit/s figures are obtained for the 'optimistic' scenario (pd=1e-8, etaD=70%), which the paper itself labels as not yet demonstrated in space, and for 20 cm apertures only the 'idealized' scenario (pd=4e-10, etaD=90%) yields positive rates. Please either present the demonstrated-technology rates as the headline or explicitly label the optimistic/idealized assumptions as future-technology projections in the abstract.","section":"Abstract and Section IV C 3"},{"comment":"Every positive key rate in the paper relies on the MMSE pre-compensation model: the paper states in Section IV C 1 that with SoA correction 'we did not obtain a positive key rate for any aperture diameter under these conditions.' The MMSE model is taken from the authors' previous work [32,62] and is not validated against an independent experiment or simulation in this manuscript. The residual covariance model in Eqs. (A14)-(A15) assumes known phase/amplitude statistics, no wavefront-sensor noise, no temporal error, and an exact point-ahead covariance; deviations from these idealizations would reduce the achieved coupling. Because the key rate is a threshold function of the tail of the PDTE, a modest overestimate of the MMSE gain (the mean coupling gain is only ~1.5 dB at 1 m, Table II) could eliminate the claimed positive rates. Please add a sensitivity analysis of the final key rates to the assumed MMSE residual variance (e.g., by scaling the covariance or adding uncorrected wavefront-sensor noise), and discuss how the results would degrade if MMSE performance fell between the SoA and ideal cases.","section":"Section IV C 1 and Appendix A"}],"minor_comments":[{"comment":"The free-space phase-drift term e^{-sigma_fs^2 Delta t^2/2} is inconsistent with the text stating sigma_fs is the per-channel drift. For two independent uplink channels, the difference Delta theta_fs,b - Delta theta_fs,a has variance 2 sigma_fs^2, so the exponent should be -sigma_fs^2 Delta t^2. As written, the phase error is underestimated by a factor of sqrt(2) in the standard deviation. The numerical impact is small for the parameters used here (sigma_fs Delta t << 1), but the formula and text should be reconciled.","section":"Appendix C, Eq. (C3)"},{"comment":"The sentence 'the linewidth effect follows a Gaussian distribution with a standard deviation of sqrt(2) sigma_nu' is ambiguous: the sqrt(2) factor arises from the difference between Alice's and Bob's laser frequencies, not from the round index. Please rephrase to clarify which difference is being considered.","section":"Text before Eq. (C3)"},{"comment":"The simulated key rates are point estimates from Monte Carlo sampling of the turbulence and jitter distributions, but the number of samples and the statistical uncertainty are not reported. Given that the 20 cm aperture cases lie near threshold, provide confidence intervals or at least state the number of Monte Carlo draws.","section":"Section IV"},{"comment":"The claim that the GEO satellite coverage is 'approximately one-third of the planet's surface' is not quantified with a source; consider adding a reference or a clarifying calculation.","section":"Section I"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is potentially interesting for the journal, but I encourage the editor to watch the novelty disclosure concerning the MMSE model: Refs. [32] and [62] are from the same research group, and the present paper's central result inherits the validity of that model without adding independent validation. The abstract should be tempered to avoid overstating the detector realism. These issues are addressable in a major revision."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Rough take: this is the first quantitative study I know applying TF-QKD and MP-QKD to a GEO untrusted-node scenario with AO pre-compensation, and it does a serious job on the channel modeling. The headline rates—about 260 bit/s for TF, 180 bit/s for MP with 1 m OGS—are new, the protocol choices make sense, and the paper is honest about the detector scenarios in the body. The abstract overstates 'realistic detectors': the few-hundred-bit/s numbers come from the optimistic case (pd = 1e-8, ηD = 70%), not currently space-demonstrated SNSPDs. With the space-demonstrated parameters (Y0 = 100 Hz, ηD = 50%) you get 17 bit/s for MP-QKD with a 1 m OGS and nothing for TF. That distinction is in the paper, so it's a presentation problem, not a hidden result.\n\nThe real soft spot is the one the authors admit: every positive rate goes through the MMSE pre-compensation model. With state-of-the-art correction they get no positive key rate for any aperture. The MMSE model comes from their own earlier simulation work; there is a wave-optics cross-check in Fig. 3, but the specific GEO uplink performance with the 18.5 µrad point-ahead angle is not independently validated. This makes the feasibility claims conditional on a simulated AO gain. I would not call that fatal—it is a modeling assumption in a performance analysis—but the abstract and conclusion should flag that the 20 cm feasibility and the hundreds-of-bit/s rates are MMSE-dependent, not current-technology statements.\n\nOn the reader's Eq. (C3) concern: I checked it and the sqrt(2) is consistent. The difference of two lasers has variance 2σν², giving the 8π²σν² term in σ_tot². So I'd write that off as a minor or non-issue.\n\nMinor things: no error bars on the key rates even though the PDTE is a distribution, and the phase drift model relies on one severe-turbulence VERTIGO scenario. Neither changes the conclusion.\n\nWho should read it: anyone working on satellite QKD architectures or AO pre-compensation for uplinks. It is a solid performance reference, not a breakthrough, but it fills a real gap. I'd send it to peer review.","headline":"First serious rate estimates for GEO untrusted-node TF/MP-QKD; results hinge on simulated MMSE gain and optimistic detector assumptions, but the paper is careful and deserves peer review.","tokens_in":19838,"tokens_out":3528,"would_cite":true,"duration_ms":36557,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"A single geostationary satellite carrying two 50 cm telescopes could support untrusted-node quantum key distribution with terrestrial stations of 20 cm to 1 m aperture, achieving secret-key rates of a few hundred bits per second with…","keywords":["quantum key distribution","twin-field QKD","mode-pairing QKD","measurement-device-independent QKD","geostationary satellite","adaptive optics uplink precompensation","free-space optical turbulence","untrusted-node quantum network"],"falsifier":"Measure the uplink coupling efficiency of a 1 m ground station with MMSE precompensation to a GEO terminal at 30° elevation and 1550 nm under turbulence comparable to the model's ($r_0=25$ cm, $\\theta_0=8.51\\ \\mu$rad). If the mean turbulence coupling falls below the simulated value of about $0.56$ or the mean one-way channel attenuation stays above roughly $60$ dB, the predicted few-hundred-bit/s rates would not survive; a free-space phase drift several times larger than the assumed $150$ rad/s would likewise break the TF-QKD phase-locking and MP-QKD pairing assumptions.","tokens_in":18840,"feed_emoji":"🛰️","tokens_out":10420,"duration_ms":102710,"temperature":0.7,"pith_summary":"Quantum key distribution over continental distances usually forces a choice between trusting intermediate relay nodes or building very large telescopes. This paper argues that a single geostationary satellite acting as an untrusted measurement node, equipped with two 50 cm telescopes and talking to ground stations with apertures from 20 cm to 1 m, can avoid both costs by using two loss-resilient protocols, twin-field and mode-pairing QKD. In the best case with realistic detectors, the simulated secret-key rates reach roughly 260 bit/s for TF-QKD and 180 bit/s for MP-QKD at 1 m ground apertures, and positive rates appear even at 20 cm once detector performance matches ground-based superconducting detectors. The result matters because such a node would serve roughly a third of the planet's surface without trusted relays and without requiring meter-class ground infrastructure.","feed_headline":"GEO quantum keys: up to 260 bit/s from a 1 m dish","feed_subtitle":"Twin-field and mode-pairing QKD plus adaptive optics make 20 cm ground telescopes viable links.","key_machinery":"The carrying mechanism is a reciprocity identity: the coupling efficiency of the precompensated uplink into the satellite receiver equals the coupling of the satellite receiver mode back-propagated to the ground and coupled to the transmitter mode, evaluated at the point-ahead angle. This turns the GEO uplink with adaptive-optics correction into a downlink problem solvable with known phase and log-amplitude statistics, with a numerical overlap integral as the final step. The paper compares a standard on-axis phase correction with an MMSE estimator that reconstructs the phase at the point-ahead angle from downlink phase and amplitude measurements. The resulting per-channel efficiency distributions are convolved with satellite-jitter losses and fixed losses to form the probability distribution of transmission efficiency $\\tau$, which is then fed into square-root-scaling secret-key-rate models for sending-or-not-sending TF-QKD and MP-QKD; a Gaussian phase-drift model with standard deviation $\\sigma_{\\mathrm{fs}}=150$ rad/s sets the phase-locking requirement for TF-QKD and the optimal maximal pairing length $L_{\\mathrm{max}}$ for MP-QKD.","core_discovery":"The paper's central claim is that a GEO satellite carrying two 50 cm telescopes can act as an untrusted measurement node for TF-QKD and MP-QKD with practical ground stations. Using the simulated channel model with MMSE adaptive-optics precompensation, the optimum secret-key rate at 1 m ground apertures is about $260$ bit/s for TF-QKD and $180$ bit/s for MP-QKD when detectors have 70% efficiency and dark-count probability $10^{-8}$; with detector parameters matching ground-based superconducting nanowire detectors (90% efficiency, dark-count probability $4\\times10^{-10}$), both protocols give positive rates down to 20 cm apertures, reaching $822$ bit/s for TF-QKD and $280$ bit/s for MP-QKD at 1 m. With currently demonstrated space-qualified detector parameters (50% efficiency, 100 Hz dark count), only MP-QKD with 1 m ground stations gives a positive rate, $17$ bit/s. The authors conclude that detector performance and advanced adaptive-optics correction, rather than telescope size, set the feasibility boundary for a scalable GEO untrusted-node QKD service.","pith_inferences":["A consequence the paper leaves implicit is network-level: the same per-link rates, combined with GEO coverage geometry, imply a single satellite could act as a key-distribution hub for many small ground stations, with total throughput set by detector dark counts and scheduling rather than by telescope size.","The detector comparison sets a concrete technology target: space-qualified SNSPDs with dark-count probability below about $10^{-8}$ and efficiency above about 70% would make the 20 cm terminal viable, a threshold that a dedicated space-demonstration mission could test before committing to a full QKD service.","The assumed free-space phase-drift standard deviation of $150$ rad/s implies TF-QKD phase locking over GEO needs only millisecond-scale feedback; an experimental measurement of the drift on a real GEO uplink would settle whether phase locking is a modest engineering task or a dominant cost."],"forward_implications":["With 1 m ground stations and MMSE precompensation, the predicted maximum rates are ~260 bit/s for TF-QKD and ~180 bit/s for MP-QKD under the optimistic realistic detector scenario (70% efficiency, dark-count probability $10^{-8}$).","Under the currently demonstrated space-detector parameters (50% efficiency, 100 Hz dark count), the model still yields 17 bit/s for MP-QKD with two 1 m stations, so a positive-rate GEO link is within reach of present hardware.","If space detectors reach ground-commercial performance (90% efficiency, 1 Hz dark count), both protocols give positive rates down to 20 cm ground apertures, with rates of 822 bit/s (TF-QKD) and 280 bit/s (MP-QKD) at 1 m.","With standard, non-MMSE adaptive-optics correction, the model finds no positive key rate at any studied aperture, making the advanced precompensation a necessary ingredient of the predicted performance.","MP-QKD achieves the same order of magnitude of key rate as TF-QKD without global phase locking, which the paper identifies as a practical advantage for space deployment."],"supporting_citations":[{"why":"Supplies the MMSE phase-estimation model and pseudo-analytical statistics that generate the precompensated uplink coupling efficiencies.","marker":"[32]"},{"why":"Provides the asymmetric sending-or-not-sending TF-QKD security model and the intensity-balancing condition used for the TF key-rate estimates.","marker":"[46]"},{"why":"Provides the asymmetric MP-QKD asymptotic key-rate model, including pairing rate and decoy-state formulas the simulation evaluates.","marker":"[20]"},{"why":"Gives the phase-drift and frequency-offset model used to compute the MP-QKD misalignment error and optimize the maximal pairing length.","marker":"[51]"},{"why":"Supplies the experimentally demonstrated space SNSPD parameters (50% efficiency, 100 Hz dark count) used for the pessimistic detector scenario.","marker":"[49]"},{"why":"Supplies the probability distribution of transmission efficiency (PDTE) convolution framework and turbulence statistics used for channel loss modeling.","marker":"[35]"},{"why":"Establishes the reciprocity principle that lets the uplink with AO precompensation be modeled as a downlink at the point-ahead angle.","marker":"[26]"},{"why":"Documents the free-space TF-QKD phase-locking demonstration that underlies the assumed residual phase error of about 0.1%.","marker":"[43]"},{"why":"Reports the ground-to-GEO adaptive-optics precompensation demonstration supporting the assumed feasibility of the uplink correction.","marker":"[23]"},{"why":"Defines the sending-or-not-sending twin-field QKD protocol whose decoy-state security bounds the TF key-rate simulation.","marker":"[14]"}],"fun_headline_variants":["GEO QKD: untrusted satellite node works with 20 cm dishes","Satellite QKD: 20 cm ground telescopes get positive key rates","Twin-field QKD via GEO: 260 bit/s from a 1 m dish","Mode-pairing QKD from GEO: feasible with 20 cm telescopes"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the advanced MMSE adaptive-optics precompensation achieves in a real GEO uplink the coupling efficiencies the simulation assigns to it; with only standard correction the model predicts no positive key rate at any studied aperture.","fun_headline_variants_meta":{"raw":{"variants":["GEO QKD: untrusted satellite node works with 20 cm dishes","Satellite QKD: 20 cm ground telescopes get positive key rates","Twin-field QKD via GEO: 260 bit/s from a 1 m dish","Mode-pairing QKD from GEO: feasible with 20 cm telescopes"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001068,"raw_usage":{"total_tokens":4501,"prompt_tokens":999,"completion_tokens":3502,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":615,"completion_tokens_details":{"reasoning_tokens":3417}},"tokens_in":615,"tokens_out":3502,"duration_ms":28194,"temperature":1.0,"reasoning_tokens":3417,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T10:43:15.232537+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the uplink coupling efficiency of a 1 m ground station with MMSE precompensation to a GEO terminal at 30° elevation and 1550 nm under turbulence comparable to the model's ($r_0=25$ cm, $\\theta_0=8.51\\ \\mu$rad). If the mean turbulence coupling falls below the simulated value of about $0.56$ or the mean one-way channel attenuation stays above roughly $60$ dB, the predicted few-hundred-bit/s rates would not survive; a free-space phase drift several times larger than the assumed $150$ rad/s would likewise break the TF-QKD phase-locking and MP-QKD pairing assumptions.","supporting_citations":[{"cited_title":"Lognon´ e, Optimization of High Data Rate Ground to Satellite Links Pre-compensated by Adaptive Optics , Ph.D","cited_arxiv_id":null,"evidence_quote":"Supplies the MMSE phase-estimation model and pseudo-analytical statistics that generate the precompensated uplink coupling efficiencies."},{"cited_title":"Zhang, W","cited_arxiv_id":null,"evidence_quote":"Provides the asymmetric sending-or-not-sending TF-QKD security model and the intensity-balancing condition used for the TF key-rate estimates."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the asymmetric MP-QKD asymptotic key-rate model, including pairing rate and decoy-state formulas the simulation evaluates."},{"cited_title":"Takenaka, A","cited_arxiv_id":null,"evidence_quote":"Gives the phase-drift and frequency-offset model used to compute the MP-QKD misalignment error and optimize the maximal pairing length."},{"cited_title":"Liu, W.-J","cited_arxiv_id":null,"evidence_quote":"Supplies the experimentally demonstrated space SNSPD parameters (50% efficiency, 100 Hz dark count) used for the pessimistic detector scenario."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the probability distribution of transmission efficiency (PDTE) convolution framework and turbulence statistics used for channel loss modeling."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Establishes the reciprocity principle that lets the uplink with AO precompensation be modeled as a downlink at the point-ahead angle."},{"cited_title":"Zhong, J","cited_arxiv_id":null,"evidence_quote":"Documents the free-space TF-QKD phase-locking demonstration that underlies the assumed residual phase error of about 0.1%."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Reports the ground-to-GEO adaptive-optics precompensation demonstration supporting the assumed feasibility of the uplink correction."},{"cited_title":"Huttner, R","cited_arxiv_id":null,"evidence_quote":"Defines the sending-or-not-sending twin-field QKD protocol whose decoy-state security bounds the TF key-rate simulation."}],"review_version":1}